deepakpanda93 commented on issue #16082: URL: https://github.com/apache/hudi/issues/16082#issuecomment-5393380536
This epic looks complete. All three sub-tasks are done: | sub-task | | status | |---|---|---| | HUDI-6507 | Java 11 runtime support | closed as completed (#17235, 2026-06-25) | | HUDI-6509 | Java 17 runtime support | closed as completed (#17236, 2026-06-25) | | HUDI-6510 | Java 17 compile time support | closed on JIRA | And the outcome has shipped. The root `pom.xml` at `release-1.2.0` declares `<java.version>11</java.version>`, and the 1.2.0 release notes carry a section *"Java 11 is the Default Build/Runtime"* - *"Hudi now builds and runs on Java 11 by default"*, with *"Spark 4.0 / 4.1 bundles additionally require Java 17+ at runtime"* and a full Spark 4.1 adapter and bundle. That is what this epic set out to certify. Two notes before closing. **JIRA is stale.** [issues.apache.org](http://issues.apache.org/) still shows HUDI-6507 and HUDI-6509 as Open, and HUDI-6506 as Open/Unresolved, even though both sub-task issues were closed as completed in June. Worth updating so JIRA-based triage does not keep resurfacing this. **One regression came out of this work and is still open: #19589.** Raising `java.version` to 11 pinned only `hudi-common`, `hudi-io` and the hadoop-mr/hadoop-common modules back to `<release>8</release>`, so the released Spark 3.x bundles contain 1,418 Java 11 classes - which contradicts the same release note's statement that *"Spark 3.x bundles continue to support Java 8+."* That is a Java 8 compatibility defect rather than a gap in Java 11/17 support, and it is tracked on its own issue, so it should not hold this epic open. Closing on that basis. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
